Slashdot Log In
Plotting/Graphing Programs for Mac OS X
Posted by
pudge
on Mon Mar 03, 2003 11:44 AM
from the he-has-a-degree-in-science dept.
from the he-has-a-degree-in-science dept.
brarrr writes "I'm starting out in graduate school at the UW in Materials Science and Engineering and doing research on spin electronics. Results from this work have me searching for a Mac OS X plotting/graphing program for 2D data and there are many of them, but no useful comparison anywhere. What do you use? What do you recommend? Why? My uses will include plotting, presentation, curve fit, trendline analysis, and more. I've looked briefly at: pro Fit, gnuplot (difficult to use, not very professional output), Abscissa (site is down, cannot evaluate), SmileLab (not very robust), Tecplot, IGOR (so far the best looking, but expensive), and KaleidaGraph (difficult to use, feels poorly ported). So what works/doesn't work? And don't bother saying Excel...."
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Full
Abbreviated
Hidden
Loading... please wait.
What I use (Score:5, Informative)
Re:What I use (Score:5, Insightful)
Parent
Re:What I use (Score:4, Informative)
Parent
Have you tried grace? (Score:5, Informative)
--JJ
Re:Have you tried grace? (Score:3, Informative)
I have never tried the OSX version, but I'm sure it'll be just as good.
Gnuplot is nice, and does 3D plotting too, but like the poster says, it doesn't make publication quality plots.
Re:Have you tried grace? (Score:1, Informative)
IBM Data Explorer (Score:5, Informative)
Excellent (and surprisingly easy) for complex 3D stuff but a bit over the top for simple 2D plots.
It's Free Software, but I don't know if it works for MacOS though.
Re:IBM Data Explorer (Score:1)
METAPOST (Score:5, Informative)
If so, use METAPOST [bell-labs.com].
On that page is a specific section on formatting graphs with METAPOST. It is a bit more complicated than throwing them out of MATLAB (which is what I'd choose for looking at the results day to day), but I guarantee that it will look more "professional", and well just plain better than any other method. And it is free. It also has the advantage of being based on Donald Knuth's work with METAFONT (itself work for TeX, which is essential for all maths/scientific publications) and it is what he uses for TAOCP. That should be enough.
Re:METAPOST (Score:1)
Well, regardless of what you use for initial viewing/analysis of results, use METAPOST for presentation.
Re:METAPOST (Score:2)
Prizm (Score:5, Informative)
Re:Prizm (Score:1)
Gnuplot (Score:4, Informative)
Re:Gnuplot (Score:4, Informative)
Once I got it working gnuplot has done everything I need. PNG output for web, eps output for embedding, ps output for ps2pdf. Don't bother with gnuplot's 'pdf' output mode. It depends on a very suckful PDF generation library (that stamps all your output with big "demo mode" watermark crap).
My dad uses IGOR and is very happy with it. Of course, his employer bought it for him.
Parent
A MacOSX version exists (Score:4, Informative)
Parent
Jgraph (Score:2, Informative)
R? (Score:4, Informative)
http://www.r-project.org (has the official Carbon build, an X11 super package with something like 300 external packages is also occasionally built, check the R-SIG-Mac mailinglist for information about that)
Re:R? (Score:2, Informative)
R is multi-platform, capable of producing graphs in a number of formats (eps, ps, pdf, png etc), can be run in batch mode (great for simulations) and can produce very pretty graphics. There is no gui, so all fine tuning of the figures has to be done through the command line. However, it can handle immense data sets, supports mathematical text (meaning you can throw a complicated equation into your graph and have it come out with publication-quailty finish) and can do some colour, three-dimensional plotting. A wonderful piece of software that has been indispensible during my Ph. D. research.
Gui and R (Was Re:R?) (Score:2)
There's also at least one person working on more interactive forms of documents (read: widgets and data in your paper controlling live graphing surfaces) which could be used to build GUIs as well. You can find info about some of this cutting edge stuff over at http://omegahat.org and I think http://obversive.sf.net
Until those arrive there's also Sweave (its in the tools package that ships), which while not really a GUI lets you combine LaTeX and R noweb style to generate some quite nice ps/pdf output (it deals with making sure the appropriate PDF or EPS files are around and whatnot).
Or RWeb which lets you build CGI based interfaces to R (complete with graphs), though performance isn't exactly stellar (R has a pretty heavy launch penalty relative to most computational tasks)
Aabel (Score:1, Interesting)
Ploticus (Score:3)
I've used it to create plots/graphs for journal publication and I've used it to create graphs on-demand for a website.
Main page: here [sourceforge.net]
Good examples: here [sourceforge.net]
IGOR (Score:5, Informative)
standard: $550
education: $395
student: $85
Although it suggests the student rate is for personal use only, I think I would try to find me a student.
Re:IGOR (Score:3, Informative)
Re:IGOR (Score:1)
Re:IGOR (Score:2)
Re:IGOR (Score:1)
I love Igor (Score:5, Informative)
The next best thing is the ability to save and restore your entire state (plots, command history, variable and array contents, scripts, notes,... well everything) in a compact single file that is crossplatform with windows and macs. finally it is very mac like in the way it works. I find this very nice. Other sci programs are good but dont act mac-like.
but my most favorite aspect is that this software is written by physicists and statistical quantum optics people. When you tell them what your are doing they sometimes actually get interested in it and will write you a special piece of code to do it.
the worst part of IGOR is that it is now so old, there are lots of options on every command rather than a consistent interface. So its becomeing a little less mac-like with time. this actually is fine for unix minded folks but confusing for many mac people. I hate reading man pages I like things that are intuitive.
Parent
Mathematica (Score:4, Informative)
The problem is often the the graph itself. Lots of pretty graphs and plots are more confusing than informative. Look into one of those clever books by Tufte. "The visual display of quantiative information" or something like that; they are good coffee table books as well.
The obvious answer (Score:4, Funny)
pro Fit (Score:2)
pro Fit is very nice and gives professional looking output, it is inexpensive (though not free), and is fairly easy to use for most tasks. It will also do curve fitting and other nice things through an easy-to-use graphical interface.
R is a statistical package which works fairly well for most plotting needs, though I have found it to be a little more obnoxious to use than pro Fit and much much more difficult to configure to get it to plot things the way you want (at least until you get familiar with the commands--it is a command line package). OTOH, it will do statistical analysis and such niceties as principle component analysis on your data.
Igor (Score:3, Informative)
Igor can be annoying at times, but the fact of the matter is that it doesn't run out of gas when you need something sophisticated.
Since you are a student I'd recommend taking the student license which from what I remember is fairly reasonable.
Kaleidagraph (Score:3, Informative)
PS: I think the recent port was decent. It doesn't crash and everything is in the same place as the older version.
Mathpad (Score:2)
gnuplot hard? (Score:1)
Still using CricketGraph (Score:2, Interesting)
Eventually, I do plan to get someting else. GraphPad Prism, as mentioned elsewhere in this topic, is probably the most likely replacement. I have also tried Kaleidagraph and ProFit and they're not bad.
I would like to use someting like Gnuplot which I know is open and will be around forever but It just isn't easy to use yet. I think the secret there would be to develop some spreadsheet macros to output data in Gnuplot's ASCII based file format but I have not had time to research this possibility and try it out.
Re:Still using CricketGraph (Score:1, Informative)
Re:Still using CricketGraph (Score:2)
Mathematica + Illustrator (Score:3, Interesting)
Try R - open source S+ (Score:3, Informative)
The syntax is somewhat tricky to learn but if you are a coder you shouldn't have much difficulty.
There's an aqua and an X11 version (through fink):
http://www.r-project.org
There is a pdf on doing graphic and plots available through their "Contributed" part of the website.
ANY that implement Tufte's suggestions? (Score:5, Interesting)
Conceptually, he suggests trying to erase every bit of ink that doesn't represent data.
More concretely, his book is replete with suggestions that produce graphs that are beautiful, easy to comprehend--and slightly unconventional.
For example, he proposes doing away with standard evenly graduated axes. Basically he suggests that each point on the graph should have an individual tick on the axis labelled with its exact value--and skip the round-numbered ticks (10, 20, 30). Of course this doesn't work if there are too many data points, but it is astonishing how many real-world graphs do NOT have too many.
Has anyone seen software on any platform that aims to implement Tufte's approach?
(Excel and friends have gone in exactly the opposite direction, of course. Tufte is vitriolic in his distaste for idiocy such as taking a single data point and illustrating it as a three-dimensional solid bar.)
Re:ANY that implement Tufte's suggestions? (Score:2)
You can still make stuff look awful if you try though, you do have a complete programming language at your disposal if you try
Graph-O-Matic is shareware (Score:1, Informative)
I use Mathematica every day now, but when I want a quick 2D plot, and want it to look good, I use Graph-O-Matic. It is inexpensive shareware, but it is still pretty flexible. I like that I don't have to remember any syntax for getting the axes to scale, etc. Almost everything is point and click with the mouse.
Another great thing about graph-o-matic is that it uses Quartz rendering and the plot are all anti-aliased. They look great. And you can grab the plot with the cursor and drag around to change views. Graph-O-Matic is very simple and very limited, but it is also very cool. Check it out!
I like to use: (Score:2)
Symptom of a wider open source problem (Score:3, Insightful)
Others... (Score:2, Informative)
The best have been mentioned. A couple not mentioned include:
I don't know if some of these are MacOS compatible or not. They are Unix compatible though.
How to COUNTER Excel? (Score:2)
Worse yet, I usually CAN get Excel to produce the graphs I need--after hours of struggling, trickery, hand-creating special data series, fussing with the fact that the charts are made of rubber and aren't any particular size--no problem unless you want a set of charts that are consistent with each other--can't get decent graduations on a logarithmic axis, etc. etc. etc.
Anyone have any tips on how to QUICKLY convince a manager that you have a credible, specific need that Excel can't handle?
Re:How to COUNTER Excel? (Score:2)
Tell him you're plotting those special numbers that Excel calls "########"
Re:Deltagraph? (Score:2, Informative)
Academic pricing is available....
Re:Math text book graphs (Score:2)